DECOMPOSITION AND MAPPING OF LOCALLY CONNECTED LAYERED NEURAL NETWORKS ON MESSAGE-PASSING MULTIPROCESSORS
作者:
TOM TOLLENAERE,
GUYA. ORBAN,
DIRK ROOSE,
期刊:
Parallel Algorithms and Applications
(Taylor Available online 1993)
卷期:
Volume 1,
issue 1
页码: 43-56
ISSN:1063-7192
年代: 1993
DOI:10.1080/10637199308915430
出版商: Taylor & Francis Group
关键词: Decomposition problem;mapping problem;neural networks;performance prediction
数据来源: Taylor
摘要:
In this paper we present an integrated model for decomposition and mapping (D&M) of Locally Connected Layered Neural Networks (LCLNs) on message-passing multiprocessors. Within the framework of this model we analyze two previously proposed D&M strategies for a particular class of LCLNs. The model is compared with the performance of a neural network simulation environment, running on a transputer array. We find that both strategies may be applicable, depending on the network size, and we can determine the problem size at which one strategy is preferred over the other. Furthermore, we find that the regularity of the communication pattern between the processors is an unexpected factor in the D&M decision.
点击下载:
PDF (241KB)
返 回